📚 Part of the OBX-319 Regulatory Dossier — Reader's Guide. This article shows the live document; edits to the source appear here automatically.

---

Module 4.2.3 — Toxicology (OBX-319)

Document ID: M4-2.3
Version: 1.0
Change History: 1.0 — Initial issue.
Standard(s): ICH M3(R2), S6(R1), S5(R3), S7A; S2/S1/S7B addressed by waiver rationale

4.2.3 Toxicology — OBX-319

This module contains the toxicology study reports supporting OBX-319, a humanised IgG1 bispecific monoclonal antibody in which one Fab arm engages CD19 and the second engages CD20, produced by recombinant Chinese hamster ovary (CHO) cell culture, purified through a Protein A capture and polishing downstream train, and administered subcutaneously for the treatment of moderate-to-severe active Systemic Lupus Erythematosus (SLE). The programme was designed under the biotechnology-derived framework of ICH S6(R1), with ICH M3(R2) governing the scope and timing of studies relative to the pivotal Phase 3 study OBX319-301 (randomised, double-blind, placebo-controlled, 1:1:1, 52 weeks; N=480 randomised — 162 OBX-319 High, 158 OBX-319 Low, 160 Placebo; baseline SLEDAI-2K approximately 11) and the marketing application, which is submitted as a BLA under 21 CFR 601. Because OBX-319 recognises human and cynomolgus monkey CD19 and CD20 but does not cross-react with rodent orthologues, the cynomolgus monkey is the sole pharmacologically relevant toxicology species; rodent studies would not be scientifically informative and were not conducted. Cell-substrate, adventitious-agent and product-/process-related impurity safety (CHO host-cell protein and DNA, aggregate and charge-variant control across the Protein A and polishing steps) is established in the Quality module under ICH Q5A(R2), Q5C and Q6B and is not reproduced here. The findings across the programme are dominated by the anticipated, on-target exaggerated pharmacology of B-cell depletion, are reversible on cessation of exposure, and reveal no target-organ toxicity independent of the intended immunopharmacology.

4.2.3.1 Nonclinical Safety Program

The program supporting OBX319-301 (randomized, double-blind, placebo-controlled) comprised:

Study type	Species / system	Duration	GLP
Repeat-dose toxicity + TK	cynomolgus monkey	up to 26 weeks	GLP
Safety pharmacology (CV/respiratory endpoints embedded)	cynomolgus monkey	within repeat-dose	GLP
Tissue cross-reactivity	human & monkey tissues	in vitro	GLP
Enhanced pre/post-natal development (ePPND)	cynomolgus monkey	per ICH S6(R1)/S5(R3)	GLP
Local tolerance	monkey	within repeat-dose	GLP

Species selection and study design. The cynomolgus monkey was qualified as the single pharmacologically relevant species on the basis of arm-specific, high-affinity binding to both cynomolgus CD19 and CD20 and confirmed absence of rodent cross-reactivity; the tissue cross-reactivity study corroborated conserved, B-lymphoid-restricted target distribution between human and monkey. The pivotal repeat-dose programme used the intended clinical subcutaneous route with weekly administration and comprised an initial dose-range-finding phase followed by the pivotal chronic study of up to 26 weeks, each incorporating a vehicle control and ascending OBX-319 dose levels together with main-study and recovery cohorts of both sexes to characterise reversibility. The high dose was selected to produce and maintain maximal (near-complete) B-cell depletion and generous systemic exposure, and the low/intermediate levels were positioned to bracket the pharmacologically active range. The first-dose level and monitoring strategy were guided by an in vitro cytokine-release evaluation and a minimum anticipated biological effect level (MABEL) rationale appropriate to a B-cell-engaging bispecific antibody. Toxicokinetics, core safety-pharmacology endpoints, local-tolerance assessment and anti-drug antibody sampling were integrated into these studies rather than conducted as standalone protocols, consistent with ICH S6(R1).

Pivotal repeat-dose findings. The principal and expected finding was on-target exaggerated pharmacology: sustained, near-complete depletion of circulating B lymphocytes accompanied by marked depletion of B-cell-dependent zones — germinal centres and follicles — in spleen, lymph nodes and gut-associated lymphoid tissue, with reduced germinal-centre cellularity. Dose-related reductions in serum immunoglobulins (hypogammaglobulinaemia, principally IgM and IgG) accompanied the depletion. T-lymphocyte, myeloid, erythroid and platelet lineages were preserved, and there was no drug-related toxicity in any organ system outside the immune compartment. Subcutaneous administration was locally well tolerated. During the treatment-free recovery phase, peripheral B cells repopulated and immunoglobulins returned toward baseline, demonstrating reversibility once exposure declined and B-cell reconstitution resumed. Because the only treatment-related findings were the anticipated, reversible, non-adverse extension of the intended pharmacology, the no-observed-adverse-effect level (NOAEL) was established at the highest dose tested. The theoretical increase in susceptibility to infection accompanying profound B-cell depletion was specifically monitored (clinical signs, body weight, and haematology/immunophenotyping).

Safety pharmacology (embedded). In accordance with ICH S6(R1) and S7A, cardiovascular endpoints (blood pressure, heart rate and quantitative electrocardiography including corrected QT interval, by telemetric or serial assessment), respiratory endpoints (rate and pattern) and central-nervous-system/neurobehavioural observations were captured within the repeat-dose studies. No adverse cardiovascular, respiratory or CNS effects attributable to OBX-319 were identified. A dedicated hERG assay and a standalone thorough-QT study were not warranted for an intact immunoglobulin (see waiver rationale).

Tissue cross-reactivity. A GLP tissue cross-reactivity study was performed by immunohistochemistry across a full panel of human tissues with a corresponding cynomolgus monkey panel. Membrane staining was confined to B-lymphocyte-containing structures consistent with the known distribution of CD19 and CD20, with no unexpected reactivity in non-lymphoid tissues. The result supports both the relevance of the monkey model and the low likelihood of off-target toxicity for a molecule whose targets are lineage-restricted.

Enhanced pre- and post-natal development (ePPND). Consistent with ICH S5(R3) and S6(R1), an ePPND study was conducted in the cynomolgus monkey. Because IgG placental transfer is greatest during the second and third trimesters, the design evaluated the potential for fetal and neonatal B-cell depletion together with maternal exposure and pregnancy outcomes. As anticipated for an effector-competent IgG1 directed at B-lineage antigens, transient depletion of infant peripheral B cells was observed, with reconstitution during the post-natal period and no teratogenic signal; lymphoid development and infant immunophenotype were assessed through the recovery interval. These findings inform pregnancy and lactation labelling and the risk-management documentation.

Local tolerance. Local tolerance at the subcutaneous injection site was evaluated within the repeat-dose studies. Injection-site reactions are an expected consequence of subcutaneous antibody administration; observed reactions were characterised for incidence, severity and reversibility to support the clinical presentation.

GLP compliance. The pivotal repeat-dose toxicity, tissue cross-reactivity, ePPND, and the embedded safety-pharmacology and local-tolerance assessments were conducted in compliance with Good Laboratory Practice. Exploratory dose-range-finding and mechanistic work was non-GLP, as is standard.

Studies not conducted (with justification)

• Genotoxicity (ICH S2) — not warranted for a large protein that does not interact directly with DNA; standard bacterial and mammalian genotoxicity assays are neither applicable nor scientifically informative for an intact recombinant immunoglobulin, and DNA-reactive process-related impurities are controlled within the Quality module (ICH Q6B) rather than through animal genotoxicity testing
• Standard rodent carcinogenicity (ICH S1) — not scientifically informative for a targeted biologic and not feasible given the absence of rodent cross-reactivity; assessed via a weight-of-evidence carcinogenicity risk assessment that considers the immunomodulatory mechanism, the theoretical impact of sustained B-cell depletion on immune surveillance, the absence of any proliferative or pre-neoplastic finding in the chronic cynomolgus study, and the accumulated clinical experience with B-cell-depleting therapeutics
• hERG / dedicated thorough-QT (ICH S7B/E14) — not applicable to a monoclonal antibody, which lacks a physicochemical basis for direct cardiac ion-channel or repolarisation effects; cardiovascular endpoints, including corrected QT, were instead covered within the monkey toxicity studies
• Standard rodent PK & metabolism — no rodent cross-reactivity, so rodent exposure would not be pharmacologically relevant; elimination proceeds by proteolytic catabolism to peptides and amino acids with FcRn-mediated recycling, and cytochrome-P450/transporter-based disposition does not apply to an intact IgG, so classical rodent ADME and mass-balance studies were omitted

4.2.3.2 Integrated NOAEL and Safety Margins

Safety margins are expressed as the ratio of systemic exposure (AUC/Cmax) at the NOAEL in the cynomolgus monkey to the projected human exposure at the clinical dose levels evaluated in the trial. Margins are derived against the clinical dose levels evaluated in OBX319-301 (OBX-319 High, OBX-319 Low, Placebo).

NOAEL basis. Because the treatment-related effects in the chronic cynomolgus study were limited to the intended, reversible pharmacology (B-cell depletion, reduced lymphoid germinal-centre cellularity, and dose-related hypogammaglobulinaemia) with no associated organ toxicity, the NOAEL was assigned to the highest dose tested. Systemic exposure at the NOAEL was defined from the integrated toxicokinetics as steady-state AUC and Cmax, accounting for the accumulation conferred by FcRn-mediated recycling over the weekly subcutaneous dosing interval.

Exposure and interspecies projection. OBX-319 exhibits target-mediated drug disposition (TMDD): clearance is non-linear at low concentrations, where the saturable B-cell target sink dominates, and approaches linearity once the target is saturated at the exposures maintained in the toxicology studies. Human exposure at the OBX-319 High and OBX-319 Low subcutaneous doses was projected using cross-species (monkey-to-human) PK/TMDD modelling anchored to the cynomolgus toxicokinetics, and anti-drug antibody status was used to interpret exposure where relevant. On this basis, the systemic-exposure margins at the cynomolgus NOAEL relative to projected clinical exposures are adequate to support the doses evaluated in OBX319-301, and maternal exposures achieved in the ePPND study likewise exceeded projected clinical exposure. Margins are interpreted in the context of the modality: the pharmacological effect that defines the therapeutic mechanism — near-complete B-cell depletion — is intended and shared between the nonclinical and clinical settings, so the relevant safety consideration is the depth, duration and reversibility of that effect and its immunological sequelae rather than a distinct toxicological endpoint.

4.2.3.3 Carcinogenicity and Class Findings

Consistent with the immunomodulatory mechanism, serious and opportunistic infections are an important identified risk, and injection-site and hypersensitivity/immunogenicity reactions are expected; there is no boxed warning for the class.

Carcinogenicity risk assessment. In the absence of a scientifically meaningful rodent bioassay, carcinogenic potential was addressed through a weight-of-evidence assessment per ICH S6(R1). The assessment integrated the restricted, B-lineage-limited target biology, the theoretical influence of sustained B-cell depletion on immune surveillance, the lack of any proliferative, hyperplastic or pre-neoplastic finding in the chronic cynomolgus study, and the extensive clinical class experience with B-cell-depleting antibodies. No carcinogenicity signal was identified that would warrant additional nonclinical study.

Class risks and risk management. The identified and potential risks are those of the pharmacological class of B-cell-depleting agents. Serious and opportunistic infections and hypogammaglobulinaemia are the key identified risks of profound B-cell depletion and warrant immunoglobulin monitoring, infection vigilance, and the risk-minimisation measures set out in the risk-management documentation. Injection-site and hypersensitivity/infusion reactions and immunogenicity (anti-drug antibodies) are expected and are managed through labelling and pharmacovigilance. No boxed warning is warranted for the class, and endpoints associated with other therapeutic classes — for example thyroid C-cell/medullary effects, which are a consideration for GLP-1 receptor agonists — are mechanistically irrelevant to a B-cell-depleting antibody and are not applicable here.

Translational corroboration. The nonclinical mechanism and its immunological sequelae are corroborated by the clinical pharmacodynamic response in OBX319-301: the active arms produced near-complete depletion of circulating CD19+ B cells (from approximately 210 to approximately 7 cells/µL) versus no meaningful change on placebo, with falling anti-dsDNA autoantibody titres and normalisation of complement C3/C4 in responders. This pharmacology tracks the dose-ordered clinical efficacy at Week 52 — low-disease-activity/SRI-4 response (SLEDAI-2K ≤ 4) of 52.4% (76/145) on OBX-319 High and 33.8% (49/145) on OBX-319 Low versus 6.0% (9/150) on Placebo, and SLEDAI-2K LS-mean change of -6.37 (High) and -5.62 (Low) versus -3.46 (Placebo), differences of -2.91 and -2.17 versus placebo — reinforcing that the class-based risk profile, and not an unexpected toxicity, governs the safety characterisation.

Immunogenicity

Anti-drug antibody (binding and neutralising) assessment is integral; immunogenicity may affect exposure and is characterised with a tiered validated assay strategy.

Assay strategy and interpretation. Anti-drug antibodies were evaluated with a validated, tiered ligand-binding approach — screening, confirmatory (specificity), and titre assays, supplemented by a neutralising-antibody assay — consistent with FDA and EMA immunogenicity guidance, enabling exposure and toxicokinetic data to be interpreted in the context of immunogenicity. In the cynomolgus monkey, anti-drug antibodies were interpreted only for their effect on exposure and toxicokinetics: where they developed, they could accelerate clearance and reduce systemic exposure, and this was taken into account in the margin assessment. Animal immunogenicity is not predictive of the clinical immunogenicity of a humanised antibody; human immunogenicity, including the impact of anti-drug antibodies on pharmacokinetics, efficacy and hypersensitivity, is characterised in the clinical modules and managed through labelling and pharmacovigilance.

Governing guidelines: ICH M3(R2), S6(R1), S5(R3), S7A; S2/S1/S7B addressed by waiver rationale.